The label escape error: determination of the active bone-forming surface in histologic sections of bone measured by tetracycline double labels.

Twenty healthy female beagles, all postbreeding age, were given two bone-labeling agents. Transilial bone biopsies were obtained and embedded undecalcified. Double-labeled, single-labeled, and unlabeled surfaces were determined along with the interlabel width on unstained sections under fluorescent light. Mean osteon thickness was determined on toluidine blue-stained sections. A mathematical model for predicting the amount of singly labeled surface was tested and compared to the amount of surface actually found occupied by single labels of either color. We found that the model accurately predicted the amount of singly labeled surface and that the amount of singly labeled surface labeled by one agent was equal to that labeled by the other (r = 0.977, P less than .001). We conclude that singly labeled surfaces in these animals were so labeled because bone formation either began or ended in the interval between administration of the two labels and that the proposed model accurately predicts this quantity of singly labeled surface without the need to measure any singly labeled surface. This label escape phenomenon must be taken into account when measuring bone formation rates by tetracycline labeling, and the model should be used whenever it is suspected that single labels appear that are not due to label escape.